The purpose of this study is to determine whether HPA activity, as measured by ACTH release, is higher in depressed than control subjects. As ACTH is subject to negative glucocorticoid feedback on the pituitary, accurate measurement of HPA activity requires removal of the negative feedback by blocking adrenal production of cortisol with metyrapone. Part I of this study will examine the circadian secretory profiles of ACTH and cortisol in depressed vs. control subjects under both masked (baseline) and unmasked (metyrapone) conditions. In part II, the unmasking strategy will be further employed to evaluate pituitary sensitivity to human corticotropin releasing factor (hCRF) in depressives vs. noraml controls. In both part I and part II, ACTH is predicted to be abnormally increased in depressives, especially when the pituitary is unmasked with metyrapone. Part I calls for 12 depressed and 12 control subjects, ages 18-75. Blood samples for cortisol and ACTH assays are obtained every 10 minutes, first under a 24 hour baseline condition, then for 24 hours when the pituitary is unmasked by the oral administration of metyrapone. Part II calls for the comparison of two circadian times of hCRF administration (corresponding to the peak and nadir of spontaneous HPA axis activity). 20 depressed and 20 control subjects, ages 18-75, are tested in each circadian phase (80 subjects total), first under masked, then unmasked conditions, with frequent blood samples for ACTH and cortisol assays obtained around the time of hCRF administration.